Key takeaways
- Semiconductor valuation multiples contract approximately 43% from June
- AMD, Arista and Astera report record quarterly revenue
- Investors question AI financing despite continued chip demand
Artificial intelligence stocks are experiencing a sharp valuation reset even as major semiconductor and networking suppliers report record sales, exposing a widening gap between Wall Street's expectations and actual AI infrastructure demand.
Bernstein analyst Stacy Rasgon found that the Philadelphia Semiconductor Index's forward price-to-earnings ratio had fallen from approximately 28 in June to 16 by late October.
That represents a 43% contraction in the earnings multiple investors are willing to pay, even though semiconductor companies continue benefiting from expanding AI investment.
Semiconductor Valuations Fall Despite Strong Earnings
The Philadelphia Semiconductor Index had declined approximately 30% from its June peak by October 30, according to Dow Jones Market Data, despite remaining roughly 50% higher for the year.
The reversal comes amid the broader AI chip stock selloff, which erased approximately $1.64 trillion in semiconductor market capitalization during six October trading sessions.
Investors increasingly question whether cloud companies can finance enormous infrastructure commitments without sacrificing cash flow or taking on excessive debt.
The impact became evident when Intel stock collapsed nearly 29% despite reporting 78% annual growth in its Data Center and AI division.
A similar divergence affected Marvell and Palantir shares, with both companies facing selling pressure despite substantial AI-related business activity.
AMD, Arista and Astera Report Record AI Growth
Corporate results present a contrasting picture.
According to AMD's November 3 earnings release, data-center revenue surged 97% year over year to $8.55 billion, while total company revenue increased 45% to $13.38 billion.
Arista Networks reported $3.66 billion in quarterly revenue, up 58.5%, marking its fastest annual growth since its 2014 IPO.
Meanwhile, Astera Labs posted $584.9 million in revenue, representing a 154% increase as demand for AI connectivity products accelerated.
These figures suggest that data-center operators continue purchasing processors, networking equipment and supporting infrastructure.
However, strong sales do not automatically justify elevated share prices.
The distinction helps explain why an earlier Nasdaq-100 correction occurred even as technology companies continued expanding their businesses.
Why Wall Street Is Repricing the AI Boom
A stock's valuation reflects both expected earnings and the multiple investors assign to those earnings.
For example, a company earning $10 per share would be valued at $280 at a 28× multiple but only $160 at 16×, assuming earnings remain unchanged.
That illustrates how shares can fall even without deteriorating profitability.
Bernstein's analysis suggests investors are becoming more cautious about construction delays, financing costs and the returns generated by AI infrastructure.
The concern extends beyond individual chipmakers.
October's broader stock-market rotation showed that 72% of S&P 500 companies gained even as major AI stocks weakened.
An earlier Broadcom valuation reassessment similarly demonstrated how analysts can reduce price targets despite projecting substantial AI revenue growth.
